Study on Mitigation of Interfacial Intermetallic Compounds by Applying Alternating Magnetic Field in Laser-Directed Energy Deposition of Ti6Al4V/AA2024 Dissimilar Materials

Abstract

Brittle intermetallic compounds (IMCs) at the interface of dissimilar materials can seriously affect the mechanical properties of the dissimilar components. Introducing external assisted fields in the fabrication of dissimilar components is a potential solution to this problem. In this study, an alternating magnetic field (AMF) was introduced for the first time in the additive manufacturing of Ti6Al4V/AA2024 dissimilar alloy components by laser-directed energy deposition (L-DED). The effect of the AMF on the interfacial IMCs’ distribution was studied. The results indicate that the contents of the IMCs were different for different magnetic flux densities and frequencies, and the lowest content was obtained with a magnetic flux density of 10 mT at a frequency of 40 Hz. When an appropriate AMF was applied, the IMC layer was no longer continuous at the interface, and the thickness was notably decreased. In addition, the influence of the AMF on the temperature distribution and fluid flow in the melt pool was analyzed through numerical simulation. The simulation results indicate that the effect of the AMF on the temperature of the melt pool was not significant, but it changed the flow pattern inside the melt pool. The two vortices inside the cross-section that formed when the AMF was applied caused different orientations of club-shaped IMCs inside the deposition layer. A sudden change in the streamline direction at the bottom of the longitudinal cross-section of the melt pool can affect the formation of the IMC layer at the interface of dissimilar materials, resulting in inconsistent thickness and even gaps. This work provides a useful guidance for regulating IMCs at dissimilar material interfaces.

1. Introduction

Dissimilar materials can not only leverage the performance advantages of both materials simultaneously but also reduce the weight of the components [1,2]. In today’s pursuit of energy conservation, emission reduction, and sustainable development, dissimilar material structures have received increasing attention. Specifically, lightweight dissimilar structures made of titanium and aluminum alloys have been widely used in the aerospace and automotive fields [3,4].

The traditional joining methods employed to manufacture titanium/aluminum dissimilar materials mainly include fusion welding processes that use a laser, electron beam, or arc as heat sources, solid-state welding processes (diffusion bonding and friction stir welding), and brazing, including transient liquid-phase (TLP) diffusion bonding [5,6,7,8,9]. In recent years, with the development of additive manufacturing (AM) technology, titanium/aluminum dissimilar materials have been fabricated by AM that show higher flexibility and economic benefits [10,11]. As of now, the main processes for the additive manufacturing of titanium/aluminum dissimilar materials include powder-based and wire-based laser-directed energy deposition (L-DED) [10,11], laser powder bed fusion (LPBF) [12], wire arc additive manufacturing (WAAM) [13], electron-beam powder bed fusion (EBPBF) [14], and ultrasonic additive manufacturing (UAM) [15]. Each of these additive manufacturing technologies has its own advantages and disadvantages. UAM has a lower energy utilization rate and is not applicable for large and complex structures. Additive manufacturing processes using wire as the raw material have a lower molding accuracy than those using powders. Despite their high forming precision, the EBPBF and LPBF technologies are limited by the size of the forming chamber, making it difficult to fabricate large-sized structures. Moreover, the recycling and reuse of powders in the manufacturing of multi-material structures is relatively difficult, resulting in serious powder waste in EBPBF and LPBF. In contrast, the powder-based L-DED technology, which does not need a chamber and can deliver multiple material powders separately, is a very good choice for manufacturing titanium/aluminum dissimilar structures.

Like in the joining of titanium and aluminum alloys, the presence of brittle intermetallic compounds (IMCs) at the titanium/aluminum interface is also a significant challenge in the AM of titanium/aluminum dissimilar structures. The presence of IMCs not only affects the mechanical properties of the dissimilar alloy components but also leads to cracking during the printing process [16]. To solve this problem, extensive research has been conducted on composition regulation, the structural optimization of joints, and the introduction of external assisted fields. Composition regulation mainly employs two methods, i.e., establishing a material gradient transition and adding an interlayer [17,18]. These methods can effectively avoid rapid changes in composition at the interface of dissimilar materials, but they can also lead to multiple IMC layers and multiple types of IMCs. Some researchers have tried to improve the mechanical properties of titanium/aluminum dissimilar materials by optimizing the joint design to reduce or even offset the adverse effects of IMCs [19,20,21]. Although certain improvements have been made in terms of mechanical properties, inevitable IMCs still seriously affect the mechanical properties of dissimilar materials. Researchers have proposed to mitigate IMCs by adding external fields when manufacturing dissimilar material structures. The external fields so far reported mainly include an ultrasonic field and a magnetic field [22,23]. As a non-contact external field, a magnetic field has wider applicability than an ultrasonic field.

Sun et al. [24] used an axial alternating magnetic field (AMF) hybrid CMT welding–brazing process to join pure titanium TA2 and the aluminum alloy 6061-T6. The results showed that the grain size of the welded and heat-affected zone was refined after adding an alternating magnetic field, and the Al-Ti IMCs were reduced. Sun et al. [25] used a high-speed camera to observe the droplet transfer process during the axial alternating magnetic field-assisted CMT welding of Al6061-T6 and TC4. The results indicated that under the action of the magnetic field, the droplets rotated around the metal wire, and the diffusion ability of the melt pool was improved. In the field of the laser welding of titanium/aluminum dissimilar materials, Chen et al. [26] first applied an external static magnetic field (SMF) to the laser welding of Al/Ti dissimilar alloys. The results showed that the Lorentz force generated by the static magnetic field suppressed horizontal convection, resulting in a decrease in the thickness of the IMCs and a 44.4% increase in the strength of the Al/Ti joint. From this, it can be seen that magnetic fields can effectively improve the quality of titanium/aluminum dissimilar material joints. However, there is no report so far about the effect of an external magnetic field when manufacturing titanium/aluminum dissimilar structures by powder-based L-DED.

In this paper, an external axial AMF was introduced, as far as the authors know, for the first time in the manufacturing of a Ti6Al4V/AA2024 dissimilar material by L-DED. The effect of the alternating magnetic field on the IMCs at the Ti6Al4V/AA2024 interface was studied. In addition, the influence of the alternating magnetic field on the temperature distribution and fluid flow of the melt pool was calculated through a three-dimensional CFD model. This study provides useful guidance to regulate the distribution of IMCs at dissimilar material interfaces by using alternating magnetic fields.

2. Materials and Method

The titanium alloy material used in the experiment was the Ti6Al4V substrate with dimensions of 150 mm × 150 mm × 4 mm. The aluminum alloy was the AA2024 powder (Henan Yuanyang Powder Technology Co., Ltd., Xinxiang, China), and its chemical composition is listed in

Table 1. Chemical composition of AA2024 powder (wt.%) (data from material vendor).

Alloy	Cu	Mg	Mn	Si	Fe	Zn	Ti	Ni	Al
AA2024	3.98	1.71	0.81	0.47	0.31	0.08	0.05	0.04	Bal.

. The chemical composition of the Ti6Al4V substrate (Hongmao Titanium Industry Co., Ltd., Dongguan, China) is listed in

Table 2. Chemical composition of Ti6Al4V substrate (wt.%) (data from material vendor).

Alloy	Al	V	Fe	C	O	N	H	Ti
Ti6Al4V	5.88	4.02	0.13	0.05	0.16	0.007	0.015	Bal.

. The morphology of the AA2024 powder is shown in Figure 1a, and it can be observed that there were many small satellite particles attached to the AA2024 powder surface. Figure 1b shows the particle size distribution statistics of the AA2024 powder, which exhibited good near-normal distribution characteristics with an average particle diameter of 92 μm. The AA2024 powder was placed in a vacuum-drying oven (−0.1 MPa) at 120 °C for 2 h and allowed to cool naturally before being taken out for use. The surface of the Ti6Al4V substrate was polished with 400 #–1200 # sandpaper and then cleaned with acetone to remove surface oxides and oil stains.

The experiment used a self-designed powder-based L-DED system, which has been described in detail in reference [10]. Figure 2a shows the device for generating an alternating magnetic field, which consisted of a coil, a waveform generator, and a power amplifier. A sine-wave signal was generated by the waveform generator. When conducting the L-DED experiments, the coil was placed directly below the substrate to generate an axial alternating magnetic field, as shown in Figure 2b. A handheld Gauss meter was used to measure the magnetic flux density of the alternating magnetic field at the surface position of the substrate, as shown in Figure 2c. Considering the non-uniform distribution of alternating magnetic fields in space, and in order to ensure the accuracy of the experimental results, the measurement position of each experiment was located at the intersection position of the upper surface of the Ti6Al4V substrate and the coil axis, as shown in Figure 2d. During the L-DED process, the AA2024 deposition layer passed through this measurement point, and all samples for subsequent cross-sectional analyses were taken from this location. The parameters used in the experiment are shown in

Table 3. Parameters of AMF-assisted L-DED.

Parameter	Value
Laser power, W	480
Scanning speed, mm/s	15
Powder feed rate, g/min	5
Carrier gas flow rate, L/min	10
Magnetic flux density, mT	0, 5, 10, 15, 20
Magnetic field frequency, Hz	10, 20, 30, 40, 50

.

The cross-sectional samples of Ti6Al4V/AA2024 dissimilar deposits were prepared using discharge wire cutting. Preliminary polishing was performed using 400 #~2000 # sandpaper, and then the samples were polished with a diamond polishing compound with a particle size of 1–2 μm. The microstructure at the interface was observed and analyzed using optical microscopy (OM), scanning electron microscopy (SEM), and energy-dispersive spectroscopy (EDS).

3. Numerical Simulation Modeling

To study the influence mechanism of an externally applied magnetic field on the formation of IMCs, a three-dimensional CFD model that coupled heat transfer, fluid flow, and a magnetic field was established. In order to save computational costs, a symmetrical model with dimensions of 10 mm × 4 mm × 2 mm was used in the calculation. The mesh model used for the analysis is shown in Figure 3. Finer meshes in the deposition region were used to improve the computational efficiency, while the region far from the center of the heat source was calculated with coarser meshes. Due to the complex working conditions of the alternating magnetic field-assisted L-DED process, some basic assumptions needed to be made for the model during the numerical simulation [27,28]:

(1)

The interaction between the powder particles and the laser beam was not considered.

(2)

The fluid flow of the melt pool was assumed to be laminar and incompressible.

(3)

The influence of the plasma and metal vapor on the melt pool was not considered, and the heat loss caused by the evaporation was ignored.

(4)

The thermal and physical properties of the powder and substrate materials were assumed to be independent of temperature.

(5)

The powder flow distribution was assumed to be a Gaussian distribution.

3.1. Governing Equations

It is necessary to consider multiple physical processes, such as heat transfer, melt pool flow, and mass transfer, when analyzing alternating magnetic field-assisted L-DED. The mass equation, momentum equation, and energy equation are formulated as follows [28,29,30]:

The mass equation:

$${\displaystyle \frac{\partial \rho }{\partial t}}+\nabla ·\left(\rho \mathit{u}\right)=0$$

(1)

The momentum equation:

$$\rho {\displaystyle \frac{\partial \mathit{u}}{\partial t}}+\rho \left(\mathit{u}·\nabla \right)\mathit{u}=\nabla ·\left[-p\mathit{I}+\mu \left(\nabla \mathit{u}+{\left(\nabla \mathit{u}\right)}^T\right)-{\displaystyle \frac{2}{3}}\mu \left(\nabla ·\mathit{u}\right)\mathit{I}\right]+F$$

(2)

The energy equation:

$$\rho C_p{\displaystyle \frac{\partial T}{\partial t}}+\rho C_p\mathit{u}·\nabla T=\nabla ·\left(k\nabla T\right)-{\displaystyle \frac{\partial H}{\partial t}}-\rho \mathit{u}·\nabla H$$

(3)

In Equation (2), source term F can be formulated as:

$$F=\rho g\left[1-\beta \left(T-T_{melt}\right)\right]-A_{m\mathit{u}sh}{\displaystyle \frac{{\left(1-f_l\right)}^2}{{f_l}^3+c_1}}\mathit{u}+F_m$$

(4)

In Equation (3), F_m is the Lorentz force generated by the external magnetic field, which can be written as [31,32]:

$$F_m=\mathit{j}\times \mathit{B}$$

(5)

where j is the current density, which can be described by the generalized Ohm’s law:

$$\mathit{j}=\sigma \left(\mathit{E}+\mathit{u}\times \mathit{B}\right)$$

(6)

3.2. Boundary Conditions

The laser heat flux at the gas/liquid interface is as follows:

$$q_l={\displaystyle \frac{2Q\eta }{\pi r^2}}\exp \left({\displaystyle \frac{-2\left({\left(x-vt\right)}^2+y^2\right)}{r^2}}\right)-h_c\left(T-T_0\right)-{\sigma }_b\epsilon \left(T^4-T_0^4\right)$$

(7)

where Q is the laser power, η is the absorption coefficient of the laser energy, r is the effective laser beam radius, σ_b is the Stefan–Boltzmann constant, ε is the emissivity, and T₀ is the ambient temperature.

The gas/liquid interface movement V_p caused by powder mass addition can be calculated by:

$${\mathit{V}}_p={\displaystyle \frac{2m_p{\eta }_p}{{\rho }_p\pi r_p^2}}exp\left({\displaystyle \frac{-2\left({\left(x-vt\right)}^2+y^2\right)}{r_p^2}}\right)\mathit{z}$$

(8)

where m_p is the mass flow rate, η_p is the powder catchment efficiency, r_p is the mass flow radius, and z is the unit vector in the z-direction.

3.3. Validation of Numerical Model

Figure 4a,b show the comparison between the numerical and experimental profiles of the cross-section of the Ti6Al4V/AA2024 dissimilar material deposits prepared by L-DED without applying a magnetic field. It can be seen that the cross-sectional profile predicted under the simulated conditions was basically consistent with the experimentally obtained one, indicating that the numerical model used in this paper was accurate. On this basis, the effect of the AMF on the temperature and fluid flow of the melt pool was calculated.

4. Results and Discussion

4.1. Macroscopic Morphology

Figure 5 shows the optical microscopy bright-field images of the cross-section of the Ti6Al4V/AA2024 dissimilar materials under different magnetic field parameters. The magnetic flux densities corresponding to Figure 5a–d are 5 mT, 10 mT, 15 mT, and 20 mT, respectively. The alternating magnetic field frequencies corresponding to each magnetic flux density from left to right (1–5) are 10 Hz, 20 Hz, 30 Hz, 40 Hz, and 50 Hz, respectively. An increase in the frequency will increase the impedance of the coil, and increasing the electric current in the coil to too high a value will cause the coil to overheat and reduce the induced magnetic field intensity. Therefore, the actual measured magnetic flux densities in Figure 5(d4,d5) during the experiment were 19 mT and 16 mT, respectively. Figure 5e shows the cross-section of the sample without applying an alternating magnetic field. In Figure 5f, W represents the width of the AA2024 deposition layer, and H represents the height. It can be observed that there were differences in the heights and widths of the AA2024 deposition layers in the cross-section of the Ti6Al4V/AA2024 dissimilar deposits when different magnetic flux densities and frequencies of magnetic fields were applied. The Image-Pro Plus software 6.0 was used to measure the height and width. No simple linear relationship existed, and no significant correlation was found between the measured values of the heights and widths and the magnetic flux density and frequency. This paper will now focus on the influence of alternating magnetic fields on the distribution of IMCs at the interface of Ti6Al4V/AA2024 dissimilar materials.

4.2. Microstructural Analysis

In order to further analyze the changes in the IMCs at the interface of the Ti6Al4V/AA2024 dissimilar materials, the samples in Figure 5 were observed in the dark-field mode, and the results are shown in Figure 6. In the dark-field mode, the darker-colored areas near the Ti6Al4V/AA2024 interface and in some AA2024 deposition layers were IMCs. These IMCs were composed of various Ti-Al IMCs, and the types of IMCs will be determined and discussed in later sections.

In Figure 6, it can be observed that the shape and distribution of the IMCs at the interface underwent significant changes after the addition of an alternating magnetic field. However, the morphologies of the IMCs in the deposits obtained with several magnetic field parameters were very different from those obtained without a magnetic field, such as b3–b5, c4–c5, etc. The most representative one shown in Figure 6(b4) (with a magnetic flux density of 10 mT with a frequency of 40 Hz) was then selected for subsequent analysis.

Figure 7a,b show the dark-field images of the optical microscopy for the Ti6Al4V/AA2024 deposits without applying an alternating magnetic field and with a magnetic flux density of 10 mT and a frequency of 40 Hz, respectively. To better analyze the impact of alternating magnetic fields on IMCs, MATLAB programming was used to process the images in the dark-field mode. Figure 7c,d are binary images of the IMCs extracted from the processed images of Figure 7a and 7b, respectively. The changes in the IMCs are more discernable in the processed figures. The IMC layer in Figure 7c, without applying an alternating magnetic field, was continuous and roughly uniformly distributed along the Ti6Al4V/AA2024 interface and has a relatively thick thickness. After applying an alternating magnetic field, it can be observed that the IMC layer at the Ti6Al4V/AA2024 interface in Figure 7d was no longer continuous, with a gap and a significant decrease in thickness. In addition, some broken IMCs were also found in the AA2024 deposition layer, which was believed to be caused by the stirring effect of the melt pool arising from the alternating magnetic field [33,34]. In addition, the area of the AA2024 deposition layer and extracted IMCs in Figure 7 were measured using the Image-Pro software.6.0 Quantitative analysis was conducted by calculating the ratio of the area of IMCs to the entire area of the AA2024 deposition layer. The results indicate that the ratio of IMCs to the deposition area of AA2024 in Figure 7a, without applying an alternating magnetic field, was 16.9%. After applying an alternating magnetic field with a density of 10 mT and a frequency of 40 Hz, the ratio decreased to 9.7%. Therefore, applying an alternating magnetic field can effectively regulate the distribution and mitigate the content of IMCs.

The IMCs at the Ti6Al4V/AA2024 interface were further analyzed using SEM and EDS. Figure 8(a1,a2) are SEM and EDS mapping distribution images of the a1 and a2 regions in Figure 7a, respectively. Among them, Figure 8(a1) shows the area with the maximum thickness of IMCs, while Figure 8(a2) shows the area with the minimum thickness of IMCs. The EDS scanning results indicate that the IMCs were mainly composed of two elements, Ti and Al, located near the AA2024 side at the Ti6Al4V/AA2024 interface. IMCs exist in two forms: as a continuous layer and club-shaped. Club-shaped IMCs were distributed on one side of the continuous IMCs and perpendicular to the continuous IMC layer. The maximum thickness at D1 of the IMCs in the sample without applying an alternating magnetic field exceeded 40 μm, while the minimum at position D2 was close to 15 μm. Figure 8(b1,b2) are the SEM and EDS scanning distribution images of the b1 and b2 regions in Figure 7b, respectively. The stirring effect of the alternating magnetic field on the melt pool caused gaps in the IMC layer at the Ti6Al4V/AA2024 interface, as shown in Figure 7d. The SEM image shown in Figure 8(b1) indicates the presence of a very thin IMC layer at the location of the gap. The thickness at D3 of the IMC layer was less than 5 μm; therefore, it was not discernable in the dark-field mode of optical microscopy and appeared discontinuous. In addition, there were also club-shaped IMCs, as shown in Figure 8(b1), but their orientations were disordered. Figure 8(b2) shows broken IMCs, mostly distributed in flakes in the AA2024 deposition layer. Club-shaped IMCs with different orientations surrounded the sheet-like IMCs. All of these results further indicate that the alternating magnetic field had a stirring effect on the melt pool, affecting the flow of the melt pool.

Figure 9a shows the distribution of elements along path L1 in Figure 8(a1). From left to right are the AA2024 layer, continuous IMC layer, and Ti6Al4V layer. Due to the presence of club-shaped IMCs, there were fluctuations in the distribution of Ti and Al elements in the AA2024 layer. There was a transition region between the Ti and Al elements at both ends of the continuous IMC layer. Figure 9b shows the element distribution along path L2 in Figure 8(b2). The results indicate that the middle region of the sheet-like IMCs was a continuous IMC layer, and the distribution of the Ti and Al elements was relatively stable. Violent fluctuations in the Ti and Al elements existed on both sides, which were caused by the club-shaped IMCs. Furthermore, the components of the IMCs at different locations were measured through point scanning, as shown in

Table 4. Element contents in different regions.

Element, at%	P1	P2	P3	P4	P5	P6	P7
Ti	68.38	72.41	67.10	48.96	67.75	68.23	71.96
Al	31.62	27.59	32.90	51.04	32.25	31.77	28.04

. The results indicate that the intermediate positions of both the continuous layer and rod-shaped IMCs were TiAl₂ (P1, P3, P5, and P6). The transition region near the Ti6Al4V side of the continuous IMC layer was TiAl (P4), while the transition region near the AA2024 layer side was TiAl₃ (P2 and P7).

4.3. Numerical Simulation Results

The temperature distributions on the top surfaces of the deposition layers are shown in Figure 10. Figure 10a shows when no AMF was applied, while Figure 10b shows when an AMF was applied, where the magnetic flux density of the AMF was 10 mT and the frequency was 40 Hz. The results indicate that the temperature at the center of the melt pool was the highest, and there was no significant change in the temperature distribution of the melt pool after applying the AMF. The maximum temperature was increased by 17 K by introducing the magnetic field. This may be because the AMF changed the flow pattern inside the melt pool, and the Lorentz force also hindered the flow of the melt pool, ultimately affecting the convection and heat dissipation of the melt pool.

In order to analyze the influence of the AMF on the internal flow of the melt pool, sections in two directions of the melt pool were extracted, as shown in Figure 11. Figure 11a shows the streamline distribution in the cross-section of the melt pool without the AMF. It can be observed that vortices were generated on the upper surfaces of both ends of the melt pool due to the Marangoni effect. The molten metal entered the interior of the melt pool from the upper surface, and the streamline distribution was regular. In contrast, after applying the AMF, two vortices appeared inside the melt pool, as shown in the blue box in Figure 11b. These two vortices indicate that the AMF had a significant impact on the flow inside the melt pool, resulting in a disordered streamline distribution, which explains the orientation change in the club-shaped IMCs in Figure 8. Figure 11c,d show the streamline distributions in the longitudinal section of the melt pool with and without an AMF, respectively. It can be observed that there was no significant change in the flow pattern inside and on the upper surface of the melt pool after applying the AMF. However, an area where the streamline changed appeared at the bottom of the melt pool, as shown in the black box in Figure 11d. A change in the flow direction at the Ti6Al4V/AA2024 interface can affect the formation of the IMC layer, resulting in inconsistent thickness and even gaps, as shown in Figure 7. Therefore, it can be concluded that the AMF can change the flow pattern inside the melt pool, thereby affecting the formation and distribution of IMCs at the interface of dissimilar materials.

The alternating magnetic field changed the formation and distribution of IMCs, with a significant decrease in thickness and content. This is very helpful for improving the mechanical properties of Ti6Al4V/AA2024 dissimilar material structures, laying a good foundation for the engineering application of Ti6Al4V/AA2024 dissimilar materials.

5. Conclusions

In this work, the influence of an AMF on the microstructure of Ti6Al4V/AA2024 dissimilar materials prepared by L-DED was studied through experiments and numerical simulation. This work can be concluded as follows:

• The application of an AMF during L-DED changed the distribution and content of IMCs. When there was no magnetic field, the IMCs were mainly continuously distributed at the Ti6Al4V/AA2024 interface; after applying an AMF with a density of 10 mT and a frequency of 40 Hz, the IMC layer became discontinuous, and some broken IMCs entered the AA2024 deposition layer. In addition, the overall thickness of the IMCs was significantly reduced by the application of the AMF. The minimum thickness of the IMC layer was decreased from 15 μm to less than 5 μm, and the content of IMCs was decreased from 16.9% to 9.7%.
• The temperature distribution of the melt pool was not significantly affected by the AMF. However, the AMF changed the flow pattern inside the melt pool. The two vortices inside the cross-section caused different orientations of the club-shaped IMCs inside the deposition layer, while the change in the flow direction at the bottom of the longitudinal section of the melt pool affected the formation of the IMC layer at the interface of the dissimilar materials, resulting in inconsistent thickness and even gaps.